Tech in Brief

Correction October 12, 2026: the cached-input row described the discount as 90% off input for both models. It is 90% for Astra ($1.00 per 1M against $10.00) and 95% for Sol ($0.10 against $2.00); the prices listed in the table were already correct. Source: OpenAI API pricing page. Updated October 10, 2026. Every price and benchmark figure below was re-checked against the sources listed at the end on that date. The October update section near the end records what changed after this piece first ran on September 30. At DevDay 2026 on September 29, OpenAI announced two upgrades to its flagship API lineup that pull in opposite directions. GPT-6.1 Sol prices its tokens at roughly one-fifth of GPT-6 Astra's rate while scoring almost identically on independent benchmarks. A new "Ultrafast" tier for Astra generates tokens up to 8 times faster in Codex, but charges 6 times more per token to do it. Neither number means much on its own -- an 8x speed claim says nothing about wheth...

NVIDIA and Microsoft used this week's push around "AI PCs" to confirm the first wave of RTX Spark laptops, landing this fall from Asus, Dell, HP, Lenovo, Microsoft Surface and MSI, with Acer and Gigabyte to follow. The headline spec is a flagship configuration with up to 128GB of unified memory, a 20-core Grace CPU, a 6,144-core Blackwell GPU, and 1 petaflop of FP4 AI compute — paired with NVIDIA's own claim that it can run large language models up to 120 billion parameters with a 1-million-token context window, fully on-device. Specs like "128GB" and "1 petaflop" don't mean much on their own. What actually matters is what fits in that memory and what doesn't, and that's a math problem, not a marketing line. A current-generation gaming laptop (illustration only — no RTX Spark laptop has been photographed or released yet). Photo: OhanaUnited, original file page , CC BY-SA 4.0 What NVIDIA actually confirmed this week Per NVIDIA...

Apple’s iPhone 18 Pro and iPhone 18 Pro Max reach stores on September 18, turning a week of launch claims into a real buying decision. The important question is not whether the new models are Apple’s fastest phones—they are—but whether their camera controls, sustained performance, battery gains, and Siri AI rollout justify at least $1,199 for your use. This guide separates confirmed specifications from marketing language and shows who should upgrade, who should wait, and what to verify before checkout. Apple Fifth Avenue in New York City. Photo by Fletcher6 , licensed under CC BY-SA 3.0 . Cropped or altered: no. What is actually shipping today Apple says the iPhone 18 Pro and Pro Max are available in black, silver, glacier, and burgundy, with 256GB, 512GB, 1TB, and 2TB storage options. U.S. pricing starts at $1,199 for iPhone 18 Pro and $1,299 for iPhone 18 Pro Max. Those are base prices before tax, carrier-plan requirements, AppleCare, accessories, or extra storage. The...

Every laptop launched this year advertises a battery life number somewhere between 15 and 27 hours. Almost none of that number survives contact with an actual workday. The gap isn't marketing fraud, exactly — it's a testing methodology gap, and once you know what the claimed number actually measures, you can convert it into something more useful: how many real workdays you get before you need a wall outlet, and what that's worth in dollars if you'd rather just buy a power bank instead. A modern work laptop (illustration / file photo). Photo: KK IN HK, Wikimedia Commons, CC BY-SA ( original file page ) The claim, and what it actually measures Manufacturer battery claims are almost universally built on video playback — typically Netflix streaming at 150 nits of screen brightness, one of the least demanding workloads a laptop can run. A screen showing a mostly-static video frame, a GPU doing minimal decode work, background apps suspended: that is the scenario behi...

A 135-minute investigation from hardware outlet Gamers Nexus, published in early September 2026, alleges that LG smart TVs and monitors collect far more household data than most owners realize — including scanning other devices on the home Wi-Fi network, tracking viewing habits across every connected source, and in some scenarios capturing microphone audio while the set appears to be off. LG disputed the core claims in an official statement on September 10, calling them "not true" while confirming, in the same statement, that several of the underlying features do exist and are active by default. This piece separates what's independently confirmed from what's disputed, and walks through the exact settings to check if you own one of these sets. An LG smart TV on display (illustration, not the specific model tested in the investigation). Photo: Maurizio Pesce, Wikimedia Commons , CC BY 2.0 . What the investigation actually found Gamers Nexus, working with indep...

AI model distillation attacks moved from an industry argument to a U.S. government cybersecurity warning on September 8, 2026. In advisory AA26-251A, the Cybersecurity and Infrastructure Security Agency describes coordinated campaigns that use large volumes of queries, shared premium accounts, proxy networks and automated evaluation to reproduce valuable behavior from U.S. frontier models. The practical lesson is not that model distillation itself is malicious. Distillation is a legitimate engineering technique. The risk appears when another party acquires a model provider's capabilities at industrial scale while evading access controls, contractual limits and detection. Illustration: server racks at NERSC; not infrastructure involved in the reported campaigns. Photo by Derrick Coetzee, via Wikimedia Commons , dedicated under CC0 1.0 . This article separates three things that are easy to blur together: what CISA says it observed, how legitimate distillation works, and ...

Photo: Robert.Harker , original file page: Wikimedia Commons , license: CC BY-SA 3.0 On September 3, 2026, OpenAI released GPT-6 Astra and, within the same announcement, president Greg Brockman told the world to "welcome to the AGI era." That is a big claim — artificial general intelligence, the theoretical point where a machine matches or exceeds human ability across essentially any cognitive task, is the kind of milestone that shows up in history books, not product changelogs. A closer look at what OpenAI actually published, and at what independent evaluators found when they checked the same benchmarks on their own terms, shows a more complicated picture than the headline suggests. What OpenAI actually shipped on September 3 GPT-6 Astra rolled out first to a limited set of enterprise customers, including participants in OpenAI's "Daybreak Access" cybersecurity testing program, according to reporting from Axios and Fortune. A broader release to ChatGP...

Google has started rolling out a Gemini-powered "remembered items" tool inside Find Hub, letting Android phones log where a passport, spare key, or other untagged object was last placed. The feature is part of Google's September 2026 Android feature drop, announced September 1, and it targets exactly the kind of everyday object that is too small or too rarely carried to justify sticking a Bluetooth tracker on it. Google Pixel handsets, the kind of Android device that runs Find Hub and Gemini. Photo by Maurizio Pesce , file photo (illustration), via Wikimedia Commons , licensed CC BY 2.0 . What Google announced In an official post on its Android blog, Google described the new capability plainly: "Soon, you'll be able to ask Gemini to remember where you placed important items, and it will save their locations in Find Hub so you can easily find them later." The example Google gives is a voice command such as "Hey Google, remember in Find Hub that I...